Reinforcement Learning papers, 2023–2026

Key findings

• Reinforcement Learning went from 1,070 papers in 2023 to 2,985 in 2026, roughly 2.8x in three years.
• NeurIPS publishes the most Reinforcement Learning work: 2,494 papers from 2023 to 2026, 36% of the four-venue total.
• The fastest-growing overlap is with Reasoning Models: papers tagged with both rose from 9 in 2023 to 437 in 2026.

Representative papers

Spread across years and venues so every edition shows up

• NeurIPS 2026\texttt{FEROM}: Frontier Endogenous Reveal-Order Marginal Policy Optimization for Masked Diffusion LMs
• ICML 2026$\text{DT}^\text{2}$: Decision-Targeted Digital Twins
• ICLR 2026$\nabla$-Reasoner: LLM Reasoning via Test-Time Gradient Descent in Textual Space
• CVPR 2026A Causal Marriage between VLM and IRM from Understanding to
• NeurIPS 2025$\texttt{G1}$: Teaching LLMs to Reason on Graphs with Reinforcement Learning
• ICML 2025A Computationally Efficient Algorithm for Infinite-Horizon Average-Reward Linear MDPs
• ICLR 2025$\phi$-Update: A Class of Policy Update Methods with Policy Convergence Guarantee
• CVPR 2025All-Optical Nonlinear Diffractive Deep Network for Ultrafast Image Denoising
• NeurIPS 2024A Critical Evaluation of AI Feedback for Aligning Large Language Models
• ICML 2024${\rm E}(3)$-Equivariant Actor-Critic Methods for Cooperative Multi-Agent Reinforcement Learning
• ICLR 2024A Black-box Approach for Non-stationary Multi-agent Reinforcement Learning
• CVPR 2024DMR: Decomposed Multi-Modality Representations for Frames and Events Fusion in Visual Reinforcement Learning
• NeurIPS 2023“Why Not Looking backward?” A Robust Two-Step Method to Automatically Terminate Bayesian Optimization
• ICML 2023A Connection between One-Step RL and Critic Regularization in Reinforcement Learning
• ICLR 2023$O(T^{-1})$ Convergence of Optimistic-Follow-the-Regularized-Leader in Two-Player Zero-Sum Markov Games
• CVPR 2023"Seeing" Electric Network Frequency From Events
• NeurIPS 2026$\varphi$TD: Distributional Reinforcement Learning using Characteristic Functions
• ICML 2026$\textit{S}$-SPPO: Semantic-Calibrated Self-Play Preference Optimization
• ICLR 2026$AutoDrive\text{-}P^3$: Unified Chain of Perception–Prediction–Planning Thought via Reinforcement Fine-Tuning
• CVPR 2026AdaptVision Efficient Vision-Language Models via Adaptive Visual Acquisition

All Reinforcement Learning papers

911 papers, page 12 of 19

551. ICLR 2026Policy Likelihood-based Query Sampling and Critic-Exploited Reset for Efficient Preference-based Reinforcement Learning
552. ICLR 2026Policy Newton Algorithm in Reproducing Kernel Hilbert Space
553. ICLR 2026PolicyFlow: Policy Optimization with Continuous Normalizing Flow in Reinforcement Learning
554. ICLR 2026Polychromic Objectives for Reinforcement Learning
555. ICLR 2026Pose-RFT: Aligning MLLMs for 3D Pose Generation via Hybrid Action Reinforcement Fine-Tuning
556. ICLR 2026Post-training Large Language Models for Diverse High-Quality Responses
557. ICLR 2026PosterCraft: Rethinking High-Quality Aesthetic Poster Generation in a Unified Framework
558. ICLR 2026Potentially Optimal Joint Actions Recognition for Cooperative Multi-Agent Reinforcement Learning
559. ICLR 2026Pragma-VL: Towards a Pragmatic Arbitration of Safety and Helpfulness in MLLMs
560. ICLR 2026Predicting LLM Output Length via Entropy-Guided Representations
561. ICLR 2026Predictive CVaR Q-learning
562. ICLR 2026Preference-based Policy Optimization from Sparse-reward Offline Dataset
563. ICLR 2026PreferThinker: Reasoning-based Personalized Image Preference Assessment
564. ICLR 2026Presenting a Paper is an Art: Self-Improvement Aesthetic Agents for Academic Presentations
565. ICLR 2026Pretrain Value, Not Reward: Decoupled Value Policy Optimization
566. ICLR 2026Primal-Dual Policy Optimization for Adversarial Linear CMDPs
567. ICLR 2026Principled Fast and Meta Knowledge Learners for Continual Reinforcement Learning
568. ICLR 2026Principled RL for Diffusion LLMs Emerges from a Sequence-Level Perspective
569. ICLR 2026PrismAudio: Decomposed Chain-of-Thought and Multi-dimensional Rewards for Video-to-Audio Generation
570. ICLR 2026PRO-MOF: Policy Optimization with Universal Atomistic Models for Controllable MOF Generation
571. ICLR 2026Probing in the Dark: State Entropy Maximization for POMDPs
572. ICLR 2026Probing to Refine: Reinforcement Distillation of LLM Reasoners via Explanatory Inversion
573. ICLR 2026Process-Verified Reinforcement Learning for Theorem Proving via Lean
574. ICLR 2026Prompt Curriculum Learning for Efficient LLM Post-Training
575. ICLR 2026Prompt-MII: Meta-Learning Instruction Induction for LLMs
576. ICLR 2026ProofOptimizer: Training Language Models to Simplify Proofs without Human Demonstrations
577. ICLR 2026PROS: Towards Compute-Efficient RLVR via Rollout Prefix Reuse
578. ICLR 2026Prosperity before Collapse: How Far Can Off-Policy RL Reach with Stale Data on LLMs?
579. ICLR 2026Proximal Supervised Fine-Tuning
580. ICLR 2026ProxyThinker: Test-Time Guidance through Small Visual Reasoners
581. ICLR 2026PSP: Prompt-Guided Self-Training Sampling Policy for Active Prompt Learning
582. ICLR 2026Q-Learning with Adjoint Matching
583. ICLR 2026Q-Learning with Fine-Grained Gap-Dependent Regret
584. ICLR 2026Q-learning with Posterior Sampling
585. ICLR 2026Q-RAG: Long Context Multi‑Step Retrieval via Value‑Based Embedder Training
586. ICLR 2026QeRL: Beyond Efficiency - Quantization-enhanced Reinforcement Learning for LLMs
587. ICLR 2026QuadGPT: Native Quadrilateral Mesh Generation with Autoregressive Models
588. ICLR 2026Quagmires in SFT-RL Post-Training: When High SFT Scores Mislead and What to Use Instead
589. ICLR 2026Quantile Advantage Estimation for Entropy-Safe Reasoning
590. ICLR 2026QuestA: Expanding Reasoning Capacity in LLMs via Question Augmentation
591. ICLR 2026QuRL: Low-Precision Reinforcement Learning for Efficient Reasoning
592. ICLR 2026QuRL: Rubrics As Judge For Open-Ended Question Answering
593. ICLR 2026R-Horizon: How Far Can Your Large Reasoning Model Really Go in Breadth and Depth?
594. ICLR 2026R-Zero: Self-Evolving Reasoning LLM from Zero Data
595. ICLR 2026R1-Code-Interpreter: LLMs Reason with Code via Supervised and Multi-stage Reinforcement Learning
596. ICLR 2026R1-Reward: Training Multimodal Reward Model Through Stable Reinforcement Learning
597. ICLR 2026R2-Dreamer: Redundancy-Reduced World Models without Decoders or Augmentation
598. ICLR 2026R2PS: Worst-Case Robust Real-Time Pursuit Strategies under Partial Observability
599. ICLR 2026R4: Nested Reasoning-Retrieval for Reward Modeling in Role-Playing Agents
600. ICLR 2026Random Policy Valuation is Enough for LLM Reasoning with Verifiable Rewards

FAQ

How many Reinforcement Learning papers were published at NeurIPS, ICLR, ICML and CVPR from 2023 to 2026?

6,885 in total: 1,070 in 2023, 1,190 in 2024, 1,640 in 2025, 2,985 in 2026.

How many Reinforcement Learning papers are at NeurIPS 2026?

NeurIPS 2026 has 760 Reinforcement Learning papers, versus 784 the year before.

How many Reinforcement Learning papers are at ICLR 2026?

ICLR 2026 has 911 Reinforcement Learning papers, versus 379 the year before.

How many Reinforcement Learning papers are at ICML 2026?

ICML 2026 has 1,021 Reinforcement Learning papers, versus 405 the year before.

How many Reinforcement Learning papers are at CVPR 2026?

CVPR 2026 has 293 Reinforcement Learning papers, versus 72 the year before.

Which conference publishes the most Reinforcement Learning papers?

NeurIPS, with 2,494 papers from 2023 to 2026.